Re: copy on write for splice() from file to pipe?

From: Ming Lei
Date: Sat Feb 11 2023 - 01:18:23 EST


On Sat, Feb 11, 2023 at 11:18:38AM +0800, Ming Lei wrote:
> On Fri, Feb 10, 2023 at 02:08:35PM -0800, Linus Torvalds wrote:
> > On Fri, Feb 10, 2023 at 1:51 PM Jens Axboe <axboe@xxxxxxxxx> wrote:
> > >
> > > Speaking of splice/io_uring, Ming posted this today:
> > >
> > > https://lore.kernel.org/io-uring/20230210153212.733006-1-ming.lei@xxxxxxxxxx/
> >
> > Ugh. Some of that is really ugly. Both 'ignore_sig' and
> > 'ack_page_consuming' just look wrong. Pure random special cases.
> >
> > And that 'ignore_sig' is particularly ugly, since the only thing that
> > sets it also sets SPLICE_F_NONBLOCK.
> >
> > And the *only* thing that actually then checks that field is
> > 'splice_from_pipe_next()', where there are exactly two
> > signal_pending() checks that it adds to, and
> >
> > (a) the first one is to protect from endless loops
> >
> > (b) the second one is irrelevant when SPLICE_F_NONBLOCK is set
> >
> > So honestly, just NAK on that series.
> >
> > I think that instead of 'ignore_sig' (which shouldn't exist), that
> > first 'signal_pending()' check in splice_from_pipe_next() should just
> > be changed into a 'fatal_signal_pending()'.
>
> Good point, here the signal is often from task_work_add() called by
> io_uring.
>
> >
> > But that 'ack_page_consuming' thing looks even more disgusting, and
> > since I'm not sure why it even exists, I don't know what it's doing
> > wrong.
>
> The motivation is for confirming that if the produced buffer can be used
> for READ or WRITE. Another way could be to add PIPE_BUF_FLAG_MAY_READ[WRITE].

BTW, I meant the added flags are source/sink private flags, which are
not used by generic pipe/splice code, just used by the actual source and
sink subsystem.

thanks,
Ming